Ecological covariates of Ascaris lumbricoides infection in schoolchildren from rural KwaZulu-Natal, South Africa

Research output: Contribution to journalJournal articleResearchpeer-review

  • Elmar Saathoff
  • Annette Olsen
  • Jane D. Kvalsvig
  • Chris C. Appleton
  • Brian Sharp
  • Immo Kleinschmidt

OBJECTIVES: To identify environmental factors that could serve to predict Ascaris lumbricoides infection patterns and thus guide control efforts in the absence of epidemiological information; to assess whether A. lumbricoides infection is positively associated with the soil clay content. METHODS: Information on A. lumbricoides infection and re-infection in a cohort of primary schoolchildren and interview data on their socioeconomic background and behaviour were combined with environmental data using a geographical information system (GIS). Multivariate models served to explore the covariation of environmental and infection patterns adjusted for possible confounders. RESULTS: Prevalence maps and spatial statistics revealed considerable spatial clustering of infection in the small study area. Logistic multivariate regression models showed strong positive associations of infection with vegetation density measured as the normalized difference vegetation index (NDVI) at baseline [odds ratio (OR) for a 10% increase: 1.82; 95% confidence interval (95% CI): 1.24-2.68; P = 0.002] and after re-infection (OR: 2.22; 95% CI: 1.71-2.87; P < 0.001). We also found a strong negative association of re-infection with the sun exposure of the soil surface as estimated from digital elevation models (OR: 0.78; 95% CI: 0.88; P < 0.001). The soil clay content was only moderately positively associated with infection and re-infection. Socioeconomic and behavioural variables, although correlated with A. lumbricoides infection, did not appear to confound the above associations in the demographically homogeneous study area. Spatial analysis of the model residuals suggested that as the models accounted for most of the spatial pattern, the model standard errors should not be affected by spatial clustering. CONCLUSION: NDVI seems to have a high potential for the prediction of A. lumbricoides infection as it was strongly associated with infection patterns in the study area. Further advantages are that NDVI information is easy to use, affordable and available with global coverage.

Original languageEnglish
JournalTropical Medicine and International Health
Volume10
Issue number5
Pages (from-to)412-422
Number of pages11
ISSN1360-2276
DOIs
Publication statusPublished - 1 May 2005
Externally publishedYes

    Research areas

  • Ascaris lumbricoides, Environment, Epidemiology, Geographical information system, Normalized difference vegetation index, Remote sensing, Soil type, Spatial statistics, Vegetation

ID: 233742746